Single-Modality Supervised Joint PET-MR Image Reconstruction

We present a new approach for deep learned joint PET-MR image reconstruction inspired by conventional synergistic methods using a joint regularizer. The maximum a posteriori expectation-maximization algorithm for PET and the Landweber algorithm for MR are unrolled and interconnected through a deep learned joint regularization step. The parameters of the joint U-Net regularizer and the respective regularization strengths are learned and shared across all the iterations. Along with introducing this framework, we propose an investigation of the impact of the loss function selection on network performance. We explored how the network performs when trained with a single or a joint-modality loss. Finally, we explored under which settings a joint reconstruction was beneficial for MR reconstruction by using various undersampling factors. The results obtained on 2D simulated data show that the joint networks outperform conventional synergistic methods and independent deep learned reconstruction methods. For PET, the network trained with only a PET loss achieves a better global reconstruction accuracy than the version trained with a weighted sum of PET and MR loss terms. More importantly, the former further improves the reconstruction of PET-specific features where MR-guided methods show their limit. Therefore, using a single-modality loss to supervise the training while still reconstructing the two modalities in parallel leads to better reconstructions and improved modality-unique lesion recovery in our proposed framework. For MR, while the same effect is observed, joint reconstruction gains only occur in the presence of highly undersampled data. Single-modality loss joint reconstruction results are also demonstrated on 3D clinical PET-MR datasets.